The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes basic shell commands including ls, grep, sort, and tail to determine the next available sequence numbers for PRDs and user stories. These operations are restricted to the local docs/tasks/ directory and serve an organizational purpose within the skill's primary function.
[PROMPT_INJECTION]: The skill demonstrates an indirect prompt injection surface by processing content from local ideation and task files.
Ingestion points: Reads external data from docs/ideation/ and docs/tasks/*.md to inform its generation process.
Boundary markers: Does not implement explicit markers or instructions to the agent to ignore embedded commands within ingested files.
Capability inventory: Possesses the ability to read and write to the file system and execute shell commands for indexing.
Sanitization: No evidence of input validation or sanitization is present for the data read from project files.

prd